I help companies identify high-impact AI opportunities, prioritize them strategically, and translate them into
scalable end-to-end solutions — backed by 5+ years of hands-on experience as a Data Scientist in cloud
architectures, NLP, computer vision, and generative AI.
AI & Data Science Consultant with 5+ years of experience delivering end-to-end ML and generative AI solutions. I combine technical expertise with strong communication and a product mindset to bridge the gap between data and business, drive AI transformation, and turn complex ideas into practical results.
Short Video Generation System / OW Media Solutions GmbH
Designed and led the development of an automated short-video generation system.
Built a scalable AWS backend (Step Functions, Lambda, S3, ECS Fargate, DynamoDB).
Developed video rendering with OpenCV & FFMPEG; ensured maintainable Python code.
Supervised a Python developer and trained the client in AI workflows.
→ Modular, extensible architecture designed to support future AI models.
→ Decreased end-to-end production time from hours to just minutes.
→ Service available online: https://vydcraft.ai/
Data Scientist and Co-Founder
99flairs
Konsumgüter und Handel
< 10 Mitarbeiter
Designed and deployed an AI-driven fashion recommender based on clothing style preferences.
Built and tested models for fashion detection and outfit style prediction (PyTorch, TensorFlow).
Improved UI with Bubble.io and applied social login; published app on Google Play & App Store.
→ Translated subjective fashion preferences into consistent, machine-interpretable embeddings.
→ You can try out a PoC version on my website: https://dataskillconsult.de/image_upload/
Data Scientist and ML Engineer
Sopra Steria
Öffentlicher Dienst
>10.000 Mitarbeiter
Led the development of a PoC for automated image captioning in a governmental image database.
Built and fine-tuned a CNN for action recognition to generate accurate, descriptive captions.
Developed a modular prototype enabling scalability and future production deployment.
Managed client communication, requirement refinement, and supervised a new team member.
→ Reduced manual effort by 98%.
Data Scientist and ML Engineer
Sopra Steria
Öffentlicher Dienst
>10.000 Mitarbeiter
Designed and implemented a self-learning facial recognition system for a 400k-image database
Built scalable data pipelines and REST APIs on SAP Data Intelligence using Python.
Selected, customized, and integrated AI models (TensorFlow, PyTorch) for facial recognition.
Deployed components in Docker and ensured seamless SAP ecosystem integration (HANA, XSA).
→ Increased system efficiency by 2000× and improved recognition accuracy by 36%.
→ Reduced manual effort by 92%.
Data Engineer
Sopra Steria
Internet und Informationstechnologie
>10.000 Mitarbeiter
In this project, I was tasked with unifying two disparate systems for tracking working hours into a single source of truth. This solution was designed to provide managers with a clearer, more accurate view of their workforce capacities and future projections, enabling them to plan and allocate human resources more effectively.
My tasks were/are:
Solution Conceptualization: Developed a comprehensive concept for integrating and harmonizing the two systems into a unified platform.
Advanced SQL Development: Designed and implemented complex SQL queries with historical data management to ensure accurate tracking over time.
Data Pipeline Development: Built efficient data pipelines to automate the extraction, transformation, and loading (ETL) of data from both systems Azure and Oracle.
Data Transformation: Applied transformation techniques to standardize and structure the data for better analysis and usability.
Data Harmonization Using NLP: Leveraged Natural Language Processing (NLP) to harmonize and align the data from both systems, ensuring consistency and reducing discrepancies.
Requirements Management: Managed stakeholder requirements to ensure the solution met business needs and addressed operational pain points.
→ Unified, consistent dashboard that improved forecasting accuracy and operational planning.
Data Scientist
Complero GmbH
Internet und Informationstechnologie
< 10 Mitarbeiter
Developed a PoC to predict outdated or incorrect customer contact data for the insurance sector.
Extracted, cleaned, and analyzed GDPR-compliant customer datasets.
Applied ML and statistical analyses to detect indicators of outdated information.
Planned customer communication.
→ ML model for identifying outdated contact data
Project Director
Complero GmbH
Internet und Informationstechnologie
< 10 Mitarbeiter
Supported development of a GDPR-compliant, self-updating contact management app.
Conducted UI/UX testing and executed LinkedIn marketing campaigns.
Led customer development interviews and represented the company at trade fairs.
Wrote target-oriented business plans and performed market research.
→ Enabled an early stop of an unprofitable product.
Head of Customer Support
Radbonus UG
Internet und Informationstechnologie
< 10 Mitarbeiter
Helped scale a fast-growing cycling rewards app featured on national TV.
Established and led the customer service department (2–5 employees).
Designed structured service processes with feedback loops for product development.
Onboarded and supervised interns; conducted market and business analyses.
→ Established a clearly documented service process with defined feedback loops.
Zertifikate
Snowflake for Devs, Data Scientists
Snowflake
2026
AWS Cloud Parctitioner
AWS
2021
Portfolio
Kontaktanfrage
Einloggen & anfragen.
Das Kontaktformular ist nur für eingeloggte Nutzer verfügbar.
